President Putin's request to Russia's upper house of parliament for a resolution that would allow him to send troops into Ukraine to protect “Russian interests” in the Crimea, has thrown Russian-speaking Twitter into a frenzy of speculation.
As the chart above shows, in the last couple of days there have been hundreds of thousands of tweets mentioning the potentially impending “war” with Ukraine. This makes sense — the Crimean crisis has been sucking up air from news of the Ukrainian “Maidan” on the RuNet. Add war to the mix, and the result is obvious.
For some Russians, Crimea is a Russian region that is only part of Ukraine through a historical mistake. These are probably the Russian speakers driving the trends for words like “independence,” “autonomy,” and “sovereignty” in the last few days.
Other parts of the RuNet (including Russian-tweeting Ukrainians) are understandably against any sort of invasion, even if it is in the trappings of a “peacekeeping mission.” These are likely the folks driving up the use of words like “provocation,” “intervention,” and “occupation.”
We felt that it might be useful to look beyond simple search frequencies, and try to get at some underlying emotions associated with the conflict. We performed sentiment analysis of just under 10,000 Russian language tweets tweeted between 5:30PM March 1 and 7:00AM March 2, which mentioned “война” (“war”). These were collected from users with more than 10 followers (to cut down on spam), and excluded retweets to capture more original tweets. We found that while there are tweets expressing both positive and negative sentiments towards war in our data, the negative sentiments expressed about the war hold an edge of almost 2-to-1, when they are mapped on a “Negative-Neutral-Positive” scale. Neutral tweets are by far the most numerous, however, making up over 53% of the population.
Negative tweets are also more strongly negative — when broken down on a more detailed scale, we can see that a large portion of negative tweets falls into the highly negative values of (-4,-3,-2) as opposed to positive tweets, which mainly fall into the weakly positive (+1).
Even then, the strongly positive tweets tend to be ironic — irony is something that sentiment analysis software isn't great at picking up. So a tweet like:
ну ахринеть теперь! супер просто! да война! тебя то нам не хватало!
well goddamn! fabulous! yes war! all we were missing is you!
gets classified as positive, which means that there are even fewer positive sentiments expressed about war. This doesn't mean there aren't truly positive sentiments being expressed. For example,
@Rogozin @holmogorov Война до победного конца!Покажем этим хохлам с палками ,как стреляет АК -12.
@Rogozin @holmogorov War to the bitter end! Lets show these “khokhly” [derogatory term for Ukrainian -A.T.] with sticks, how and AK-12 works.
Negative tweets appear to be more correctly classified. Here are some examples that were classified as strongly negative (-4)
А ваще, это же пиздец, ребят! Война в Украине? Путин, ты ебанулся? Я все понимаю, амбиции и все такое, но подумай о жизнях миллионов людей!
This is f*cked up guys! War in Ukraine? Putin, have you gone off your f*cking rocker? I mean I understand, ambitions and all that, but think of the lives of millions of people!
Господи.. война – это всегда очень страшно и грозит серьезными последствиями..
Jesus.. war is always very scary and has very serious consequences..
но в первую очередь, конечно, не хочется, чтобы была война или кто-то погиб
in the first place, of course I don't want for there to be war and for anyone to die
Unfortunately, taken on the whole, this does not bode well for anti-war efforts, at least on Twitter. Although there is a sizable and vocal minority of users who appear to be opposed to war, the large majority of users seems to be ambivalent about it. In this situation a strong political leader like President Putin can pretty much do whatever he wants. Especially when the West cares much more about the Oscars than Crimea: